*************************************
**********     ROADMAP     **********
*************************************

The data and code included in this package are used to produce the analysis reported in "Exposing the Revolving Door in Executive Branch Agencies" by Logan Emery and Mara Faccio, Journal of Financial and Quantitative Analysis (accepted 2024).  The code files are labeled corresponding to the order in which they should be run (e.g., 00_ should be run first, followed by 01_, etc.).  The following is a list of the code files provided, the required program, and their function:

00_1_Contracts_NameMatch: SAS, clean procurement contract data in preparation for matching to Capital IQ data
00_2_Contracts_NameMatch: Stata, match company names from procurement contract data to Capital IQ
00_3_Contracts_Clean: SAS, clean procurement contract data in preparation for analysis
01_RegData_Analysis: SAS, produce dataset for Table 1, Table 2, and Table 3
02_1_Summary_Stats: Stata, report summary stats for Table 1, produce data for Figures 1 and 2
02_2_Connections_ByInd_NAICS: Matlab, create Figure 1
02_3_Connections_ByState: Matlab, create Figure 2
03_1_Analysis_RegData: Stata, create Table 3
03_2_RegPlots: Matlab, create Figure 3
04_Contracts_Analysis: SAS, produce dataset for Tables 4, 5 and 8 
05_1_Analysis_Contracts: Stata, create Tables 4 and 5, additional summary stats for Table 1
05_2_ContractPlots: Matlab, create Figure 4
06_Contracts_Analysis_Public: SAS, produce dataset for Tables 6 and 7
07_Analysis_Contracts_Public: Stata, create Table 6
08_1_Analysis_Contracts_Public_NNmatch: Stata, run propensity score matching for Table 7
08_2_Analysis_Contracts_Public_NNmatch_Reg: Stata, create Table 7
08_3_PSMPlots: Matlab, create Figure 5
09_Reneg_Analysis: SAS, produce dataset for Table 9
10_Analysis_Reneg: Stata, create Table 9

The code makes use of several datasets that are not provided and must be obtained by the researcher.  These include:

BoardEx_ListingMatch: list of changes in firm identifiers resulting from IPOs or privatizations, obtained via request from BoardEx
BoardEx_Coverage_20220520: list of firms receiving full coverage by BoardEx and first date of full coverage, obtained via request from BoardEx
BoardEx_Emp: director profile employment data from BoardEx (na_dir_profile_emp on WRDS)
BoardEx_Comp: company profile details data from BoardEx (na_company_profile_details on WRDS)
BX_CIQ_Merge_NAICS5: merged firm identifiers from BoardEx and Capital IQ
CIQ_NAICS: NAICS codes obtained from Capital IQ
Compustat_2022: Compustat Fundamentals Annual file
GovContracts_2000_2020: data on procurement contracts obtained from usaspending.gov
Data from RegData (4-digit-industry-summary, 5-digit, 6-digit, metadata, restrictions): see https://www.quantgov.org

The provided datasets can be found in Input Data\Open, and include:

Agency_BX_Map: mapping of US agencies to BoardEx identifiers
BoardEx_PlumBook_Agencies: mapping of US agencies as listed in the Plum Book to BoardEx identifiers
Plum_Book_2016: scanned version of the 2016 Plum Book

*************************************
**********  PRIMARY DATA   **********
*************************************

The primary dataset for the analysis, N_FormerRegulators.dta, is provided in Input Data\Open.  It contains firm-agency-year level data of the number of individuals with prior work experience at the given agency (i.e., former regulators) appointed to a position within the given firm in the given year.  The reported agency identifiers correspond to agency identifiers in RegData, and can be converted to BoardEx identifiers using the Agency_BX_Map.csv file provided in Input Data\Open.  There are two reported firm identifiers.  The first, CompanyID, corresponds to firm identifiers in BoardEx.  The second, GVKEY_Link_ID, is an identifier used to link firms to GVKEYs from Compustat.  The GVKEY linking file can be provided upon request, conditional on S&P Global's verification of the license/subscription, and approval.  Please submit your request, along with your affiliation (e.g., University), to mfaccio@purdue.edu.

Former regulators that joined the given firm within two years of leaving the agency are labeled as direct transitions (n_fr_direct) and former regulators that joined the given firm more than two years after leaving the agency are labeled as indirect transitions (n_fr_indirect).  These count variables increase in the year the former regulator is appointed and decrease the year after a former regulator departs.  For example, if a former regulator is appointed in 2014 and departs in 2015, the former regulator will be counted towards the firm's totals in 2014 and 2015.

*************************************
**********    CITATION     **********
*************************************

If you make use of the data provided in this package, please cite:

Emery, L.; and M. Faccio. "Exposing the Revolving Door in Executive Branch Agencies." Journal of Financial and Quantitative Analysis, forthcoming (2024).